The principal goal of this project is to develop novel multivalent antibody constructs having increased affinities over conventional bivalent antibodies that will better inhibit HIV infectivity. If it proves possible to make such constructs they may play an important role in situations where passive antibody therapy is warranted, including the vertical transmission of HIV. The working hypothesis is that the interaction between CD4 and gp120 is too strong to be overcome at reasonable concentrations of bivalent antibodies that have nM binding constants. A critical aspect of this hypothesis is that the CD4 - gp120 interaction at the cell surface is multivalent due to there being multiple copies of CD4 on the cell surface and multiple copies of gp120 on the virus surface. Accordingly novel antibody constructs will be developed that are comprised of an array of Fab fragments on a polymer backbone to increase the avidity of the antibody - gp120 interaction. The research will proceed in three phases. First, using conventional methods, a panel of monoclonal antibodies will be selected for maximal affinity for the peptide corresponding to positions 397 - 430 of gp120. This peptide has been implicated as a site critical to the CD4 - gp120 interaction by Laskey et. al. Cell 50,975, 1987. Second both intact antibodies and Fab fragments from the best antibody identified in phase 1 will be conjugated to linear or branched dextran polymers or to linear polyacrylamide polymers of varying lengths to form "n-valent" conjugates. These polymers will be physiochemically characterized and evaluated for their abilities to decrease the binding of gp120 to CD4 and for their abilities to inhibit HIV infectivity in vitro. Third, those conjugates demonstrating increased affinities and increased neutralizing abilities will then be used for clinical trials in situations where passive antibody therapy is warranted. Because the third phase depends on success in the first two phases funding is only requested for the first two phases.